It is a not-uncommonly-held position among those who care about such things that books in their current form factor are not long for this world. It is true that production costs for e-books will be substantially lower than for the print kind, though total cost of acquisition/editing/paginating the copy will be unchanged. I'm not an insider in the publication business, but I have to bet that the actual cost of buying the paper, running the presses, binding the books, and shipping them to wholesalers and retailers is really not the lion's share of the end cost of the book when compared to the cost of acquiring the intellectual property in the first place, editing it, and transforming it into an attractive and easily-read format. Given that, e-books will likely not be able to be sold for a price too very much lower than the current cost of paper books, at least not until they are old enough that they would, under the current system, be out of print. (Old books remaining available is probably what I would consider to be the biggest advantage of more widespread prevalence of e-books.) So people who are saying that one of the major advantages of e-books will be lower prices . . . well, they will probably be somewhat lower. But they won't be $1 new releases, that's for sure.
And that's before one takes into account the effect that book piracy will have on the cost of books, too. Though I'm not sure what percentage of potential book pirates in this hypothetical scenario are people who would have bought books under the current system at all, or perhaps are currently in the "used bookstores and libraries only" market segment. While library usage does have some effect on the market, used bookstore traffic is, so far as I know, entirely under the radar as far as market tracking is concerned, and certainly doesn't provide money upstream to the publisher and author. Buying a book that someone else is finished with is certainly not illegal, but it does have some characteristics in common with e-book piracy in that respect. Though obviously the ability to give a book away and yet also still have it is a definite factor in favor of the scale of e-book piracy...
If paper books become obsolete and e-books the norm, though, the used book market will certainly collapse, and if that happens without all the old out-of-print books being made available again in e-book format, it will cause me actual pain in my soul to think of all those books I never found that now are all gone, pulped and turned into copies of USA Today, if not post-consumer-recycled toilet paper. Rare-book stores will still exist, of course, and there will probably remain very occasional places that specialize in one form or another of old-style paper books, but it would definitely turn into a niche market at best.
All that being said, I hope the Glorious E-Book Only Future never comes to pass. Don't get me wrong, the idea of taking a Kindle on vacation with all the books I could possibly manage to read while away squirreled away in its little memory chip is very appealing. And as mentioned above, if publishers put their backlists into e-book format and release them, bibliophiles everywhere would squee in unison at the prospect of getting our grubby little hands on some of the things we've heard about but have never been able to find. But while I have no trouble seeing the appeal of the digital, the physical books have an importance to me entirely out of proportion with the importance of the information they contain. For me, the process of reading a book is a multi-sensory experience, and on-screen reading falls far short of the "real thing", as far as I'm concerned.
First, and most obviously, there's a book's appearance. They do say never to judge a book by its cover (and, OK, one appeal of a Kindle-like device would certainly be the ability not to be judged BY OTHER PEOPLE on the basis of the cover of the book one is reading), but we all do it, even those of us who read SF and know without doubt that cover art is almost uniformly godawful and rarely has any commonality at all with the contents of the book it presents. Still, it does tell us something - mostly, it tells us what the marketing department wants us to think about the book, which is a useful piece of information in itself, I think. Another visual impact of physical books, though I know this is likely to appeal only to a fairly miniscule fraction of all book customers, is the effect created by all the bookcases, packed full of books, that fill our house. It means absolutely anyone who comes to the door knows in a split second that we are book people.
Tactilely speaking, I almost never open a book for the first time without holding it in both hands for at least a few seconds, taking its measure in some subconscious fashion. There is the question of the quality of the book itself, even if it's a mass-market paperback. Does it feel light for its size, suggesting they used cheap paper? Does the ink come off on your fingers at the slightest touch? Those factors also give clues to what the publisher thinks of the book, and how much (or in those examples, how little) they cared about having it put its best foot forward. If it feels heavy for its size, and the paper looks and feels fine-grained, and the glue seems substantial and likely to last more than a handful of years before letting the pages go, and the ink is at least somewhat resistant to smudging, it's likely that they thought rather better of it than in the former case. Also, I have a personal preference for matte covers over glossy in paperbacks, and if the book I've just picked up has that particular sort of satiny feel, it's likely I will sit and savor it for a just a bit longer before opening it than I would a glossier cover. Also, while this is likely an imaginary effect, with an old book, you can almost feel the weight of the years it has seen seeping back out into the surroundings, creating an aura of antiquity around itself, and making the process of holding it into a very real bridge with another time.
And then there's the smell. I have to say, I'm with Giles on this point - books should be smelly. I love the smell of a brand-new release, of paper and ink and glue all bundled into something that my olfactory center recognizes as "new book" and which sparks a level of anticipation of which I am usually not even aware. Used books have a different but no less affecting smell - the smell of slightly oxidized pages, or the very slight mustiness that even the best-tended books pick up after a couple of decades - which years of prospecting in used bookstores have led me to associate with a sense of possibility, of potential. Some books acquired from used bookstores are less pleasant in their olfactory messages - the ones that smell like stale smoke are particularly icky - but for the most part, I love the way books smell.
One doesn't generally think of books as particularly noisy, but the little crackle a hardcover makes when it's opened for the very first time is certainly something of which I am aware, and the occasional rustle of turning pages when my husband and I are reading on the couch in the evening is a comfortable sound - it sounds like home.
I can't complete the set, I'm afraid. While I'm certain that, like every baby and toddler I have ever met, I have in fact eaten my share of books, I have never ingested one since my brain developed sufficiently to be able to lay down long-term memories and call them up again later, and so cannot comment on how books taste. My love of books may be multi-sensory, but I see no problem with leaving that one out.
So, while I can think of several circumstances in which e-books would be more convenient than the paper versions, I must say I hope the markets continue to co-exist. If they don't, we'll just have to invest in a print-on-demand machine, I guess, or else learn how to bind our own books.
